Find research datasets worth reusing

Search datasets from major research repositories and use ShareScore to quickly assess how well each record supports discovery, access, and reuse.

2,474

datasets available to search

ShareScore release 0.7.1

Dataset results

2,474 results for “Segmentation”

Learn how ShareScore rates datasets ↗

zenodo48/100

The role of injection method on residual trapping at the pore-scale in continuum-scale samples: segmented data

The experiments in this work explore the role of a variable injection rate on gas saturation and residual trapping. There are 2 experiments in this work H2L (high to low injection rate) and L2H (low to high injection rate). The workflow for processing the micro-CT images to get the segmented images is described in [1]. The following scans are included in this repository NB. all data for this repository is segmented micro-CT data.: <ol><li>Dry scan prior to experiment = merged_binning_2_38_1927</li><li>H2L during high flow  = merged_segmented_flow_09_h2lh_merged</li><li>H2L during low flow  = merged_segmented_flow_11_h2ll_2_merged</li><li>H2L at the end of drainage (no flow) =merged_segmented_flow_16_dra1_pd5_merged</li><li>H2L at the end of imbibition (no flow) =merged_segmented_flow_21_imb1_pi1_merged</li><li>L2H during low flow = merged_segmented_flow_29_2_l2hl_merged</li><li>L2H during high flow = merged_segmented_flow_30_l2hh_merged</li><li>L2H at the end of drainage (no flow) =merged_segmented_flow_31_dra2_pd1_merged</li><li>L2H at the end of imbibition (no flow)  =merged_segmented_flow_33_imb2_pi1_merged</li></ol>

opencc-by-4.0Nov 2023View details →

zenodo48/100

3D Data Derivatives of Grotta di Fumane: GigaMesh-processed, Annotations and Segmentations

Overview: This repository contains derivatives of the Open Access publication by Falcucci & Peresani [FP22]. Our derived dataset (n = 62) is used to demonstrate our segmentation algorithm [BHM23], as shown in [BLM22], [BLM23], [LBM23], and will serve as a benchmark dataset for future analyses. To date, and to the best of our knowledge, our dataset is the first dataset of annotated lithic artifacts. In addition to the annotated dataset, we will also provide the segmented [BLM23] and GigaMesh preprocessed datasets [Mar+10; MK13] (n = 732) in separate folders.  Repository description:  A detailed description of the data can be found in 3D_Data_Derivatives_of_GdF_overview.pdf. For information on the archaeological interpretation of the artifacts, please refer to the original data publication by Falcucci and Peresani (2022). In our publications, we have expanded the CSV file from Falcucci and Peresani (2022) to document the use of the extended dataset: <ul> <li> Annotated: All artifacts that are annotated are marked with a 1. </li> <li> GT_PLY: All artifacts that are annotated and included in this publication are referenced by their respective file, such as 31_gt_labels.ply. </li> <li> Bullenkamp_et_al_2022: Artifacts utilized in [BLM22] are marked with a 1 . </li> <li> Bullenkamp_et_al_2023: Artifacts utilized in [BLM23] are marked with a 1. </li> <li> Linsel_et_al_2023: Artifacts utilized in [LBM23] are marked with a 1. </li> </ul>

opencc-by-4.0Aug 2023View details →

zenodo48/100

Histological Dataset for Microvascular Segmentation of Tissue-Engineered Vascular Grafts

Objectives: The pursuit of understanding vascular tissue regeneration within tissue-engineered vascular grafts (TEVGs) is of paramount importance due to the critical role these grafts play in replacing damaged or diseased blood vessels. TEVGs offer a promising alternative to traditional grafts, with the potential to integrate into the host's tissue and support the natural regenerative processes. However, challenges such as thrombosis, inflammation, and the need for grafts that can adapt to the dynamic biological environment remain. By studying the regenerative processes in TEVGs, researchers can gain insights into the mechanisms that underpin successful graft integration and function, which is essential for improving patient outcomes in vascular surgeries. This dataset, with its detailed annotations of histological features, provides a valuable resource for developing and refining machine-learning models that can analyze and predict patterns of vascular tissue regeneration. The ability to accurately segment and quantify microvessels and immune cells in regenerated arteries is a significant step forward in distinguishing between physiological and pathological regeneration, ultimately contributing to the design of more effective and reliable TEVGs for clinical use. Ethical Approval: Experimental strategy of the study is described in detail in <a href="https://www.mdpi.com/2073-4360/14/23/5149" target="_blank" rel="noopener">[1]</a> and <a href="https://www.mdpi.com/1422-0067/24/10/8540" target="_blank" rel="noopener">[2]</a>. The study was conducted according to the guidelines of the Declaration of Helsinki, and was approved by the Local Ethical Committee of the Research Institute for Complex Issues of Cardiovascular Diseases (Kemerovo, Russia, protocol code 2020/06, date of approval: 19 February 2020). Animal experiments were performed in accordance with the European Convention for the Protection of Vertebrate Animals (Strasbourg, 1986) and Directive 2010/63/EU of the European Parliament on the protection of animals used for scientific purposes. For the implantation, we used female Edilbay sheep of 42–45 kg body weight which were received from the Animal Core Facility of the Research Institute for Complex Issues of Cardiovascular Diseases (Kemerovo, Russia) and selected for the surgery by Doppler ultrasonography to identify those having carotid artery diameter of 4.0 ± 0.2 mm. Description: The dataset comprises a collection of Whole Slide Images (WSIs) obtained from biodegradable TEVGs implanted into the carotid arteries of 20 sheep. A total of 104 WSIs were acquired, each measuring an average size of 135,000 x 123,000 pixels. These WSIs were stained using Hematoxylin and Eosin (H&E), a common practice for highlighting the structure of tissue sections, which facilitates the detailed examination of histological features. These WSIs were automatically sliced into 99,831 patches of 3,000 x 3,000 pixels and subsequently filtered, resulting in 1,401 selected patches for manual annotation. Annotation Method: Two pathologists independently selected and meticulously annotated the 1401 patches, identifying nine distinct histological features associated with vascular tissue regeneration. These features include arteriole lumen (AL), arteriole media (AM), arteriole adventitia (AA), venule lumen (VL), venule wall (VW), capillary lumen (CL), capillary wall (CW), immune cells (IC), and nerve trunks (NT). The annotations were performed using binary masks, delineating each feature within the patches. Subsequently, a senior pathologist conducted a triple verification process, reviewing and refining the annotations to ensure accuracy and consistency. The annotations are provided in the form of binary masks, meticulously defined for each feature within the patches. Dataset Split: Given the limited number of subjects studied, comprising 20 sheep, we employed a 5-fold cross-validation technique to split our dataset. This method was chosen because it allows for the efficient use of limited data, ensuring that each observation has the opportunity to be used in both the training and testing sets, thus reducing bias and providing a more accurate estimate of the model's performance. In this approach, each fold involved 16 sheep for training and the remaining 4 for testing (see Table 1 and Figure 3). This partitioning scheme was consistently applied to maintain the integrity of subject groups within each subset and to prevent data leakage. The 5-fold cross-validation is particularly beneficial for our study's objectives as it maximizes the training data available for developing robust machine learning models while also ensuring that the models are tested on unseen data, thereby enhancing the generalizability of our findings. Access to the Study: Further information about this study, including curated source code, dataset details, and trained models, can be accessed through the following repositories: <ul> <li>Source code: <a href="https://github.com/ViacheslavDanilov/histology_segmentation" target="_blank" rel="noopener">https://github.com/ViacheslavDanilov/histology_segmentation</a></li> <li>Dataset: <a href="https://doi.org/10.5281/zenodo.10838384" target="_blank" rel="noopener">https://doi.org/10.5281/zenodo.10838384</a></li> <li>Models: <a href="https://doi.org/10.5281/zenodo.10838431" target="_blank" rel="noopener">https://doi.org/10.5281/zenodo.10838431</a></li> </ul> <div> </div> <div>Table 1. Patch and feature distributions across folds and subsets <table> <tbody> <tr> <td> Fold </td> <td> Subset </td> <td> Patches </td> <td> AL </td> <td> AM </td> <td> AA </td> <td> VL </td> <td> VW </td> <td> CL </td> <td> CW </td> <td> IC </td> <td> NT </td> <td> Total </td> </tr> <tr> <td> 1 </td> <td> Train </td> <td> 1168 </td> <td> 510 </td> <td> 512 </td> <td> 220 </td> <td> 675 </td> <td> 648 </td> <td> 770 </td> <td> 765 </td> <td> 409 </td> <td> 448 </td> <td> 4957 </td> </tr> <tr> <td>1</td> <td> Test </td> <td> 233 </td> <td> 81 </td> <td> 84 </td> <td> 36 </td> <td> 186 </td> <td> 169 </td> <td> 178 </td> <td> 182 </td> <td> 91 </td> <td> 25 </td> <td> 1032 </td> </tr> <tr> <td> 2 </td> <td> Train </td> <td> 1053 </td> <td> 406 </td> <td> 411 </td> <td> 179 </td> <td> 678 </td> <td> 638 </td> <td> 743 </td> <td> 746 </td> <td> 423 </td> <td> 315 </td> <td> 4539 </td> </tr> <tr> <td>2</td> <td> Test </td> <td> 348 </td> <td> 185 </td> <td> 185 </td> <td> 77 </td> <td> 183 </td> <td> 179 </td> <td> 205 </td> <td> 201 </td> <td> 77 </td> <td> 158 </td> <td> 1450 </td> </tr> <tr> <td> 3 </td> <td> Train </td> <td> 1127 </td> <td> 507 </td> <td> 511 </td> <td> 222 </td> <td> 743 </td> <td> 702 </td> <td> 759 </td> <td> 760 </td> <td> 299 </td> <td> 423 </td> <td> 4926 </td> </tr> <tr> <td>3</td> <td> Test </td> <td> 274 </td> <td> 84 </td> <td> 85 </td> <td> 34 </td> <td> 118 </td> <td> 115 </td> <td> 189 </td> <td> 187 </td> <td> 201 </td> <td> 50 </td> <td> 1063 </td> </tr> <tr> <td> 4 </td> <td> Train </td> <td> 1064 </td> <td> 466 </td> <td> 472 </td> <td> 199 </td> <td> 611 </td> <td> 566 </td> <td> 759 </td> <td> 758 </td> <td> 423 </td> <td> 291 </td> <td> 4545 </td> </tr> <tr> <td>4</td> <td> Test </td> <td> 337 </td> <td> 125 </td> <td> 124 </td> <td> 57 </td> <td> 250 </td> <td> 251 </td> <td> 189 </td> <td> 189 </td> <td> 77 </td> <td> 182 </td> <td> 1444 </td> </tr> <tr> <td> 5 </td> <td> Train </td> <td> 1192 </td> <td> 475 </td> <td> 478 </td> <td> 204 </td> <td> 737 </td> <td> 714 </td> <td> 761 </td> <td> 759 </td> <td> 446 </td> <td> 415 </td> <td> 4989 </td> </tr> <tr> <td>5</td> <td> Test </td> <td> 209 </td> <td> 116 </td> <td> 118 </td> <td> 52 </td> <td> 124 </td> <td> 103 </td> <td> 187 </td> <td> 188 </td> <td> 54 </td> <td> 58 </td> <td> 1000 </td> </tr> </tbody> </table> </div>

opencc-by-4.0Mar 2024View details →

zenodo48/100

FISBe: A real-world benchmark dataset for instance segmentation of long-range thin filamentous structures

<h2>General</h2> For more details and the most up-to-date information please consult our project page: <a href="https://kainmueller-lab.github.io/fisbe" target="_blank" rel="noopener">https://kainmueller-lab.github.io/fisbe</a>. <h2>Summary</h2> <ul> <li>A new dataset for neuron instance segmentation in 3d multicolor light microscopy data of fruit fly brains <ul> <li>30 completely labeled (segmented) images</li> <li>71 partly labeled images</li> <li>altogether comprising &sim;600 expert-labeled neuron instances (labeling a single neuron takes between 30-60 min on average, yet a difficult one can take up to 4 hours)</li> </ul> </li> <li>To the best of our knowledge, the first real-world benchmark dataset for instance segmentation of long thin filamentous objects</li> <li>A set of metrics and a novel ranking score for respective meaningful method benchmarking</li> <li>An evaluation of three baseline methods in terms of the above metrics and score</li> </ul> <h2>Abstract</h2> Instance segmentation of neurons in volumetric light microscopy images of nervous systems enables groundbreaking research in neuroscience by facilitating joint functional and morphological analyses of neural circuits at cellular resolution. Yet said multi-neuron light microscopy data exhibits extremely challenging properties for the task of instance segmentation: Individual neurons have long-ranging, thin filamentous and widely branching morphologies, multiple neurons are tightly inter-weaved, and partial volume effects, uneven illumination and noise inherent to light microscopy severely impede local disentangling as well as long-range tracing of individual neurons. These properties reflect a current key challenge in machine learning research, namely to effectively capture long-range dependencies in the data. While respective methodological research is buzzing, to date methods are typically benchmarked on synthetic datasets. To address this gap, we release the FlyLight Instance Segmentation Benchmark (FISBe) dataset, the first publicly available multi-neuron light microscopy dataset with pixel-wise annotations. In addition, we define a set of instance segmentation metrics for benchmarking that we designed to be meaningful with regard to downstream analyses. Lastly, we provide three baselines to kick off a competition that we envision to both advance the field of machine learning regarding methodology for capturing long-range data dependencies, and facilitate scientific discovery in basic neuroscience. <h2>Dataset documentation:</h2> We provide a detailed documentation of our dataset, following the <a href="https://arxiv.org/abs/1803.09010" target="_blank" rel="noopener">Datasheet for Datasets</a> questionnaire: >> <a href="https://kainmueller-lab.github.io/fisbe/datasheet" target="_blank" rel="noopener">FISBe Datasheet</a> Our dataset originates from the <a href="https://www.janelia.org/project-team/flylight" target="_blank" rel="noopener">FlyLight project</a>, where the authors released a large image collection of nervous systems of ~74,000 flies, <a href="https://gen1mcfo.janelia.org/cgi-bin/gen1mcfo.cgi" target="_blank" rel="noopener">available for download</a> under CC BY 4.0 license. <h2>Files</h2> <ul> <li>fisbe_v1.0_{completely,partly}.zip <ul> <li>contains the image and ground truth segmentation data; there is one zarr file per sample, see below for more information on how to access zarr files.</li> </ul> </li> <li>fisbe_v1.0_mips.zip <ul> <li>maximum intensity projections of all samples, for convenience.</li> </ul> </li> <li>sample_list_per_split.txt <ul> <li>a simple list of all samples and the subset they are in, for convenience.</li> </ul> </li> <li>view_data.py <ul> <li>a simple python script to visualize samples, see below for more information on how to use it.</li> </ul> </li> <li>dim_neurons_val_and_test_sets.json <ul> <li>a list of instance ids per sample that are considered to be of low intensity/dim; can be used for extended evaluation.</li> </ul> </li> <li>Readme.md <ul> <li>general information</li> </ul> </li> </ul> <h2>How to work with the image files</h2> Each sample consists of a single 3d MCFO image of neurons of the fruit fly. For each image, we provide a pixel-wise instance segmentation for all separable neurons. Each sample is stored as a separate zarr file (<a href="https://zarr.readthedocs.io" target="_blank" rel="noopener">zarr</a> is a file storage format for chunked, compressed, N-dimensional arrays based on an open-source specification."). The image data ("raw") and the segmentation ("gt_instances") are stored as two arrays within a single zarr file. The segmentation mask for each neuron is stored in a separate channel. The order of dimensions is CZYX. We recommend to work in a virtual environment, e.g., by using conda: <code>conda create -y -n flylight-env -c conda-forge python=3.9</code> <code>conda activate flylight-env</code> <h3>How to open zarr files</h3> <ol> <li>Install the python zarr package:  <pre><code>pip install zarr</code></pre> </li> <li>Opened a zarr file with: <code>import zarr</code> <code>raw = zarr.open(<path_to_zarr>, mode='r', path="volumes/raw")</code> <code>seg = zarr.open(<path_to_zarr>, mode='r', path="volumes/gt_instances")</code> <code># optional:</code> <code>import numpy as np</code> <code>raw_np = np.array(raw)</code> </li> </ol> Zarr arrays are read lazily on-demand. Many functions that expect numpy arrays also work with zarr arrays. Optionally, the arrays can also explicitly be converted to numpy arrays. <h3>How to view zarr image files</h3> We recommend to use <a href="https://napari.org" target="_blank" rel="noopener">napari</a> to view the image data. <ol> <li>Install napari:  <pre><code>pip install "napari[all]"</code></pre> </li> <li>Save the following Python script:  <code>import zarr, sys, napari</code> <code>raw = zarr.load(sys.argv[1], mode='r', path="volumes/raw")</code> <code>gts = zarr.load(sys.argv[1], mode='r', path="volumes/gt_instances")</code> <code>viewer = napari.Viewer(ndisplay=3)</code> <code>for idx, gt in enumerate(gts):</code> <code>  viewer.add_labels(</code> <code>    gt, rendering='translucent', blending='additive', name=f'gt_{idx}')</code> <code>viewer.add_image(raw[0], colormap="red", name='raw_r', blending='additive')</code> <code>viewer.add_image(raw[1], colormap="green",  name='raw_g', blending='additive')</code> <code>viewer.add_image(raw[2], colormap="blue",  name='raw_b', blending='additive')</code> <code>napari.run()</code> </li> <li>Execute:  <pre><code>python view_data.py <path-to-file>/R9F03-20181030_62_B5.zarr</code></pre> </li> </ol> <h2>Metrics</h2> <ul> <li>S: Average of avF1 and C</li> <li>avF1: Average F1 Score</li> <li>C: Average ground truth coverage</li> <li>clDice_TP: Average true positives clDice</li> <li>FS: Number of false splits</li> <li>FM: Number of false merges</li> <li>tp: Relative number of true positives</li> </ul> For more information on our selected metrics and formal definitions please see <a href="https://arxiv.org/abs/2404.00130" target="_blank" rel="noopener">our paper</a>. <h2>Baseline</h2> To showcase the FISBe dataset together with our selection of metrics, we provide evaluation results for three baseline methods, namely <a href="https://github.com/Kainmueller-Lab/PatchPerPix" target="_blank" rel="noopener">PatchPerPix (ppp)</a>, <a href="https://github.com/google/ffn" target="_blank" rel="noopener">Flood Filling Networks (FFN)</a> and a non-learnt application-specific <a href="https://www.biorxiv.org/content/10.1101/2020.06.07.138941v1" target="_blank" rel="noopener">color clustering from Duan et al.</a>. For detailed information on the methods and the quantitative results please see <a href="https://arxiv.org/abs/2404.00130" target="_blank" rel="noopener">our paper</a>. <h2>License</h2> The FlyLight Instance Segmentation Benchmark (FISBe) dataset is licensed under the <a href="https://creativecommons.org/licenses/by/4.0" target="_blank" rel="noopener">Creative Commons Attribution 4.0 International (CC BY 4.0) license</a>. <h2>Citation</h2> If you use FISBe in your research, please use the following BibTeX entry:  <pre><code>@misc{mais2024fisbe, title = {FISBe: A real-world benchmark dataset for instance segmentation of long-range thin filamentous structures}, author = {Lisa Mais and Peter Hirsch and Claire Managan and Ramya Kandarpa and Josef Lorenz Rumberger and Annika Reinke and Lena Maier-Hein and Gudrun Ihrke and Dagmar Kainmueller}, year = 2024, eprint = {2404.00130}, archivePrefix ={arXiv}, primaryClass = {cs.CV} }</code></pre> <h2>Acknowledgments</h2> We thank Aljoscha Nern for providing unpublished MCFO images as well as Geoffrey W. Meissner and the entire FlyLight Project Team for valuable discussions. P.H., L.M. and D.K. were supported by the HHMI Janelia Visiting Scientist Program. This work was co-funded by Helmholtz Imaging. <h2>Changelog</h2> There have been no changes to the dataset so far. All future change will be listed <a href="https://kainmueller-lab.github.io/fisbe/changelog" target="_blank" rel="noopener">on the changelog page</a>. <h2>Contributing</h2> If you would like to contribute, have encountered any issues or have any suggestions, please <a href="https://github.com/Kainmueller-Lab/fisbe/issues" target="_blank" rel="noopener">open an issue</a> for the FISBe dataset in the accompanying github repository. All contributions are welcome!

opencc-by-4.0Apr 2024View details →

zenodo48/100

Ultrafast imaging recordings from the axon initial segment of neocortical layer-5 pyramidal neurons.

This dataset contains imaging and whole-cell electrophysiological recordings from neocortical layer-5 pyramidal neuron from brain slices of the mouse. Electrophysiological recordings (at 20 kHz) are from the soma. Imaging data (10 kHz) are from lines along the axon initial segment (distal>proximal) with 500 nm pixel resolution. These correspond to: <ul> <li>Sodium imaging (Figures 1 and S6).</li> <li>Voltage imaging (Figures 2,4,5,S4,S7)</li> <li>Calcium imaging (Figures 3,S3,S8).</li> </ul> This dataset is used in the paper available online: Filipis L, Blömer LA, Montnach J, De Waard M, Canepari M. Nav1.2 and BK channels interaction shapes the action potential in the axon initial segment. bioRxiv, 2022. doi: 10.1101/2022.04.12.488116.

opencc-by-4.0Jan 2022View details →

zenodo48/100

S1S2-Water: A global dataset for semantic segmentation of water bodies from Sentinel-1 and Sentinel-2 satellite images

The S1S2-Water dataset is a global reference dataset for training, validation and testing of convolutional neural networks for semantic segmentation of surface water bodies in publicly available Sentinel-1 and Sentinel-2 satellite images. The dataset consists of 65 triplets of Sentinel-1 and Sentinel-2 images with quality checked binary water mask. Samples are drawn globally on the basis of the Sentinel-2 tile-grid (100 x 100 km) under consideration of pre-dominant landcover and availability of water bodies. Each sample is complemented with metadata and Digital Elevation Model (DEM) raster from the Copernicus DEM.This work was supported by the German Federal Ministry of Education and Research (BMBF) through the project "Künstliche Intelligenz zur Analyse von Erdbeobachtungs- und Internetdaten zur Entscheidungsunterstützung im Katastrophenfall" (AIFER) under Grant 13N15525, and by the Helmholtz Artificial Intelligence Cooperation Unit through the project "AI for Near Real Time Satellite-based Flood Response" (AI4FLOOD) under Grant ZT-IPF-5-39.

opencc-by-4.0Sep 2023View details →

zenodo48/100

SubPipe: A Submarine Pipeline Inspection Dataset for Segmentation and Visual-inertial Localization

<h1>Abstract</h1> This paper presents SubPipe, an underwater dataset for SLAM, object detection, and image segmentation.  SubPipe has been recorded using a lightweight autonomous underwater vehicle (LAUV), operated by OceanScan MST, and carrying a sensor suite including two cameras, a side-scan sonar, and an inertial navigation system, among other sensors. The AUV has been deployed in a pipeline inspection environment with a submarine pipe partially covered by sand. The AUV's pose ground truth is estimated from the navigation sensors. The side-scan sonar and RGB images include object detection and segmentation annotations, respectively. State-of-the-art segmentation, object detection, and SLAM methods are benchmarked on SubPipe to demonstrate the dataset's challenges and opportunities for leveraging computer vision algorithms. To the authors' knowledge, this is the first annotated underwater dataset providing a real pipeline inspection scenario. The dataset and experiments are publicly available <a href="https://github.com/remaro-network/SubPipe-dataset">online.</a> On Zenodo we provide three versions for SubPipe. One is the full version (SubPipe.zip, ~80GB unzipped) and two subsamples: SubPipeMini.zip, ~12GB unzipped and SubPipeMini2.zip, ~16GB unzipped. Both subsamples are only parts of the entire dataset (SubPipe.zip). SubPipeMini is a subset, containing semantic segmentation data, and it has interesting camera data of the underwater pipeline. On the other hand, SubPipeMini2 is mainly focused on underwater side-scan sonar images of the seabed including ground truth object detection bounding boxes of the pipeline. For (re-)using/publishing SubPipe, please include the following copyright text: SubPipe is a public dataset of a submarine outfall pipeline, property of Oceanscan-MST. This dataset was acquired with a Light Autonomous Underwater Vehicle by Oceanscan-MST, within the scope of Challenge Camp 1 of the H2020 <a href="https://remaro.eu/">REMARO</a> project. More information about OceanScan-MST can be found at <a href="https://www.oceanscan-mst.com/">this link</a>. <h1>Cam0 — GoPro Hero 10</h1> <h4>Camera parameters:</h4> <ul> <li>Resolution: 1520×2704</li> <li>fx = 1612.36</li> <li>fy = 1622.56</li> <li>cx = 1365.43</li> <li>cy = 741.27</li> <li>k1,k2, p1, p2 = [−0.247, 0.0869, −0.006, 0.001]</li> </ul> <h1>Side-scan Sonars</h1> Each sonar image was created after 20 “ping” (after every 20 new lines) which corresponds to approx. ~1 image / second. Regarding the object detection annotations, we provide both COCO and YOLO formats for each annotation. A single COCO annotation file is provided per each chunk and per each frequency (low frequency vs. high frequency), whereas the YOLO annotations are provided for each SSS image file. Metadata about the side-scan sonar images contained in this dataset: <table> <tbody> <tr> <td>Images for object detection</td> <td> </td> </tr> <tr> <td># Low Frequency (LF):</td> <td>  5000</td> </tr> <tr> <td>LF image size:</td> <td>2500 × 500</td> </tr> <tr> <td># High Frequency (HF):</td> <td>  5030</td> </tr> <tr> <td>HF Image size</td> <td>5000 × 500</td> </tr> <tr> <td>Total number of images:</td> <td>10030</td> </tr> <tr> <td>Annotations </td> <td> </td> </tr> <tr> <td># Low Frequency:</td> <td>  3163</td> </tr> <tr> <td># High Frequency:</td> <td>  3172</td> </tr> <tr> <td>Total number of annotations:</td> <td>  6335</td> </tr> </tbody> </table>

opencc-by-4.0Oct 2023View details →

zenodo48/100

Comparison and practical review of segmentation approaches for label-free microscopy

This dataset contains microscopic images of PNT1A cell line captured by multiple microcopic without use of any labeling and a manually annotated ground truth for subsequent use in segmentation algorithms. Dataset also includes images reconstructed according to the methods described below in order to ease further segmentation.  See Vicar et al. Cell segmentation methods for label-free contrast microscopy: review and comprehensive comparison. BMC Bioinformatics (2019) 20:360. DOI <a href="https://doi.org/10.1186/s12859-019-2880-8">10.1186/s12859-019-2880-8</a> Code using this dataset is available at <a href="https://github.com/tomasvicar/Cell-segmentation-methods-comparison">https://github.com/tomasvicar/Cell-segmentation-methods-comparison</a> Materials and methods  Cells were cultured in RPMI-1640 medium supplemented with antibiotics (penicillin 100 U/ml and streptomycin 0.1 mg/ml) with 10% fetal bovine serum. Prior microscopy acquisition, cells were maintained at 37 cenigrade in a humidified incubator with 5% CO2. Intentionally, high passage number of cells was used (>30) in order to describe distinct morphological heterogeneity of cells (rounded and spindle-shaped, relatively small to large polyploid cells). For acquisition purposes, cells were cultivated in Flow chambers µ-Slide I Luer Family (Ibidi, Martinsried, Germany). Quantitative phase imaging (QPI) microscopy was performed on Tescan Q-PHASE (Tescan, Brno, Czech republic), with objective Nikon CFI Plan Fluor 10x/0.30 captured by Ximea MR4021MC (Ximea, Münster, Germany). Imaging is based on the original concept of coherence-controlled holographic microscope \cite{Kolman:10,Slaby:13}, images are shown in grayscale with units of pg/µm2. DIC microscopy was performed on microscope Nikon A1R (Nikon, Tokyo, Japan), with objective Nikon CFI Plan Apo VC 20x/0.75 captured by CCD camera Jenoptik ProgRes MF (Jenoptik, Jena, Germany).  HMC microscopy was performed on microscope Olympus IX71 (Olympus, Tokyo, Japan), with objective Olympus CplanFL N 10x/0.3 RC1 captured by CCD camera Hamamatsu Photonics ORCA-R2 (Hamamatsu Photonics K.K., Hamamatsu, Japan). PC microscopy was performed on a Nikon Eclipse TS100-F microscope, with a Nikon CFI Achro ADL 10x/0.25 objective captured by CCD camera Jenoptik ProgRes MF. Folder structure and file and filename description folder "source data+groundtruth" - includes raw microscopic data    (uncompressed 16-bit for DIC, HMC and PC, 32-bit for QPI) - includes manualy annotated groundtruth (zip file - imageJ ROI file, 1bit png mask) e.g.  DIC_01_raw.tif DIC_01_groundtruth_imagejROI.zip DIC_01_groundtruth_mask.png folder "reconstructions" includes reconstructed images using reconstructions with highest dice coefficient achieved.  for DIC and HMC: rDIC-Koos, rDIC-Yin, and rWeka for PC: rPC-Top-Hat, rDIC-Yin, and rWeka for QPI: rWeka note that for rWeka images numbered 01 for DIC, HMC and PC and 01-03 for QPI were used for learning. Abbreviations DIC, differential image contrast HMC, Hoffman modulation contrast PC, phase contrast QPI, quantitative phase imaging rDIC-Koos, DIC/HMC image reconstruction according to Koos et al, Sci Rep. 2016;6:30420 rDIC-Yin, DIC/HMC image reconstruction according to Yin et al, Inf Process Med Imaging. 2011;22:384-97. rPC-Yin, PC image reconstruction according to Yin et al,  Med Im Anal. 2012; 16(5):1047 rPC-Top-Hat, Top-Hat filter according to Dewan et al, IEEE Transactions on Biomedical Circuits and Systems.2014;8(5):716-728 rWeka, probability map using Trainable Weka segmentation according to Arganda-Carreras et al. Bioinformatics. 2017

opencc-by-4.0May 2018View details →

zenodo48/100

Dataset Comparison of MRI-based automated segmentation methods and functional neurosurgery targeting with direct visualization of the Ventro-intermediate thalamic nucleus at 7T

Scientific Reports - Nature - DOI : 10.1038/s41598-018-37825-8 ################################## "Comparison of MRI-based automated segmentation methods and functional neurosurgery targeting with direct visualization of the Ventro-intermediate thalamic nucleus at 7T" ################################## E. Najdenovska*, C. Tuleasca*, J. Jorge, P. Maeder, J.P. Marques, T. Roine,  D. Gallichan, J.-P. Thiran, M. Levivier, and M. Bach Cuadra *Equally contributed authors Copyright (c) - All rights reserved. University of Lausanne. 2018. To reproduce the analyses presented in the referred study, in this repository you could find the MR images acquired from nine young healthy subjects (YS1-YS5), four elderly healthy subject (ES1-ES4) and two drug-resistant tremor patients treated treated with Vim radiosurgery by Gamma Knife (P1 and P2). The provided dataset includes the following NifTI files: - MPRRAGE @3T - DWI @3T (together with the corresponding bvals and bvecs) - MP2RAGE @7T - SWI @7T - binary masks of the manual delineation of both left and right Vim respectively that were done on the SWI (as NifTI files as well). Additionally, for the young cohort (YS1-YS5) we include as well the images used for building the quadrilateral of Guiot: - T2-w @3T - T2 CISS @3T For the patients (P1 and P2), a follow-up MPRAGE (acquired at 3T) with Gadolinium enhancement is also provided. —————————————— Notes: 1. For YS3 MP2RAGE at 7T is missing, instead MPRAGE at 3T was used 2. The code performing the thalamic nuclei clustering could be found in Zenodo (DOI: 10.5281/zenodo.123768)

opencc-by-sa-4.0May 2018View details →

zenodo48/100

Image segmentations produced by BAMF under the AIMI Annotations initiative

The Imaging Data Commons (IDC)(<a href="https://imaging.datacommons.cancer.gov/">https://imaging.datacommons.cancer.gov/</a>) [1] connects researchers with publicly available cancer imaging data, often linked with other types of cancer data. Many of the collections have limited annotations due to the expense and effort required to create these manually. The increased capabilities of AI analysis of radiology images provide an opportunity to augment existing IDC collections with new annotation data. To further this goal, we trained several nnUNet [2] based models for a variety of radiology segmentation tasks from public datasets and used them to generate segmentations for IDC collections. To validate the model's performance, roughly 10% of the AI predictions were assigned to a validation set. For this set, a board-certified radiologist graded the quality of AI predictions on a Likert scale. If they did not 'strongly agree' with the AI output, the reviewer corrected the segmentation.  This record provides the AI segmentations, Manually corrected segmentations, and Manual scores for the inspected IDC Collection images. Only 10% of the AI-derived annotations provided in this dataset are verified by expert radiologists . More details, on model training and annotations are provided within the associated manuscript to ensure transparency and reproducibility.   This work was done in two stages. Versions 1.x of this record were from the first stage. Versions 2.x added additional records. In the Version 1.x collections, a medical student (non-expert) reviewed all the AI predictions and rated them on a 5-point Likert Scale, for any AI predictions in the validation set that they did not 'strongly agree' with, the non-expert provided corrected segmentations. This non-expert was not utilized for the Version 2.x additional records.   <h3>Likert Score Definition:</h3> Guidelines for reviewers to grade the quality of AI segmentations. <ul> <li>5 Strongly Agree - Use-as-is (i.e., clinically acceptable, and could be used for treatment without change)</li> <li>4 Agree - Minor edits that are not necessary. Stylistic differences, but not clinically important. The current segmentation is acceptable</li> <li>3 Neither agree nor disagree - Minor edits that are necessary. Minor edits are those that the review judges can be made in less time than starting from scratch or are expected to have minimal effect on treatment outcome</li> <li>2 Disagree - Major edits. This category indicates that the necessary edit is required to ensure correctness, and sufficiently significant that user would prefer to start from the scratch</li> <li>1 Strongly disagree - Unusable. This category indicates that the quality of the automatic annotations is so bad that they are unusable.</li> </ul>   <h3>Zip File Folder Structure</h3> Each zip file in the collection correlates to a specific segmentation task. The common folder structure is <ul> <li>ai-segmentations-dcm This directory contains the AI model predictions in DICOM-SEG format for all analyzed IDC collection files</li> <li>qa-segmentations-dcm This directory contains manual corrected segmentation files, based on the AI prediction, in DICOM-SEG format. Only a fraction, ~10%, of the AI predictions were corrected. Corrections were performed by radiologist (rad*) and non-experts (ne*)</li> <li>qa-results.csv CSV file linking the study/series UIDs with the ai segmentation file, radiologist corrected segmentation file, radiologist ratings of AI performance.</li> </ul>   <h3>qa-results.csv Columns</h3> The qa-results.csv file contains metadata about the segmentations, their related IDC case image, as well as the Likert ratings and comments by the reviewers. <div> <table> <tbody> <tr> <td> Column </td> <td> Description </td> </tr> <tr> <td> Collection </td> <td> The name of the IDC collection for this case </td> </tr> <tr> <td> PatientID </td> <td> PatientID in DICOM metadata of scan. Also called Case ID in the IDC </td> </tr> <tr> <td> StudyInstanceUID </td> <td> StudyInstanceUID in the DICOM metadata of the scan </td> </tr> <tr> <td> SeriesInstanceUID </td> <td> SeriesInstanceUID in the DICOM metadata of the scan </td> </tr> <tr> <td> Validation </td> <td> true/false if this scan was manually reviewed </td> </tr> <tr> <td> Reviewer </td> <td> Coded ID of the reviewer. Radiologist IDs start with ‘rad’ non-expect IDs start with ‘ne’ </td> </tr> <tr> <td> AimiProjectYear </td> <td> 2023 or 2024, This work was split over two years. The main methodology difference between the two is that in 2023, a non-expert also reviewed the AI output, but a non-expert was not utilized in 2024. </td> </tr> <tr> <td> AISegmentation </td> <td> The filename of the AI prediction file in DICOM-seg format. This file is in the ai-segmentations-dcm folder. </td> </tr> <tr> <td> CorrectedSegmentation </td> <td> The filename of the reviewer-corrected prediction file in DICOM-seg format. This file is in the qa-segmentations-dcm folder. If the reviewer strongly agreed with the AI for all segments, they did not provide any correction file. </td> </tr> <tr> <td> Was the AI predicted ROIs accurate? </td> <td> This column appears one for each segment in the task for images from AimiProjectYear 2023. The reviewer rates segmentation quality on a Likert scale. In tasks that have multiple labels in the output, there is only one rating to cover them all. </td> </tr> <tr> <td> Was the AI predicted {SEGMENT_NAME} label accurate?   </td> <td> This column appears one for each segment in the task for images from AimiProjectYear 2024. The reviewer rates each segment for its quality on a Likert scale. </td> </tr> <tr> <td> Do you have any comments about the AI predicted ROIs?   </td> <td> Open ended question for the reviewer </td> </tr> <tr> <td> Do you have any comments about the findings from the study scans? </td> <td> Open ended question for the reviewer </td> </tr> </tbody> </table> </div>   <h3>File Overview</h3> <h4>brain-mr.zip</h4> <ul> <li>Segment Description: brain tumor regions: necrosis, edema, enhancing</li> <li>IDC Collection: <a href="https://www.cancerimagingarchive.net/collection/upenn-gbm/">UPENN-GBM</a></li> <li>Links: <a href="../records/11582627">model weights</a>, <a href="https://github.com/bamf-health/aimi-brain-mr">github</a></li> </ul> <h4>breast-fdg-pet-ct.zip</h4> <ul> <li>Segment Description: FDG-avid lesions in breast from FDG PET/CT scans QIN-Breast</li> <li>IDC Collection: <a href="https://www.cancerimagingarchive.net/collection/qin-breast/">QIN-Breast</a></li> <li>Links: <a href="https://doi.org/10.5281/zenodo.8290054">model weights, </a><a href="https://github.com/bamf-health/aimi-breast-pet-ct">github</a></li> </ul> <h4>breast-mr.zip</h4> <ul> <li>Segment Description: Breast, Fibroglandular tissue, structural tumor</li> <li>IDC Collection: <a href="https://www.cancerimagingarchive.net/collection/duke-breast-cancer-mri/">duke-breast-cancer-mri</a></li> <li>Links: <a href="../records/11998679">model weights</a>, <a href="https://github.com/bamf-health/aimi-breast-mr">github</a></li> </ul> <h4>kidney-ct.zip</h4> <ul> <li>Segment Description: Kidney, Tumor, and Cysts from contrast enhanced CT scans</li> <li>IDS Collection: <a href="https://www.cancerimagingarchive.net/collection/tcga-kirc/">TCGA-KIRC,</a> <a href="https://www.cancerimagingarchive.net/collection/tcga-kirp/">TCGA-KIRP</a>, <a href="https://www.cancerimagingarchive.net/collection/tcga-kich/">TCGA-KICH</a>, <a href="https://www.cancerimagingarchive.net/collection/cptac-ccrcc/">CPTAC-CCRCC</a></li> <li>Links: <a href="https://doi.org/10.5281/zenodo.8277845">model weights, </a><a href="https://github.com/bamf-health/aimi-kidney-ct">github</a></li> </ul> <h4>liver-ct.zip</h4> <ul> <li>Segment Description: Liver from CT scans</li> <li>IDC Collection: <a href="https://www.cancerimagingarchive.net/collection/TCGA-LIHC/">TCGA-LIHC</a></li> <li>Links: <a href="https://doi.org/10.5281/zenodo.8270230">model weights, </a><a href="https://github.com/bamf-health/aimi-liver-ct">github</a></li> </ul> <h4>liver2-ct.zip</h4> <ul> <li>Segment Description: Liver and Lesions from CT scans</li> <li>IDC Collection: <a href="https://www.cancerimagingarchive.net/collection/hcc-tace-seg/">HCC-TACE-SEG</a>, <a href="https://www.cancerimagingarchive.net/collection/colorectal-liver-metastases/">COLORECTAL-LIVER-METASTASES</a></li> <li>Links: <a href="../records/11582728">model weights</a>, <a href="https://github.com/bamf-health/aimi-liver-tumor-ct">github</a></li> </ul> <h4>liver-mr.zip</h4> <ul> <li>Segment Description: Liver from T1 MRI scans</li> <li>IDC Collection: <a href="https://www.cancerimagingarchive.net/collection/TCGA-LIHC/">TCGA-LIHC</a></li> <li>Links: <a href="https://doi.org/10.5281/zenodo.8290123">model weights, </a><a href="https://github.com/bamf-health/aimi-liver-mr">github</a></li> </ul> <h4>lung-ct.zip</h4> <ul> <li>Segment Description: Lung and Nodules (3mm-30mm) from CT scans</li> <li>IDC Collections: <ul> <li><a href="https://www.cancerimagingarchive.net/collection/anti-pd-1_lung/">Anti-PD-1-Lung</a></li> <li><a href="https://www.cancerimagingarchive.net/collection/lung-pet-ct-dx/">LUNG-PET-CT-Dx</a></li> <li><a href="https://www.cancerimagingarchive.net/collection/nsclc-radiogenomics/">NSCLC Radiogenomics</a></li> <li><a href="https://www.cancerimagingarchive.net/collection/rider-lung-pet-ct/">RIDER Lung PET-CT</a></li> <li><a href="https://www.cancerimagingarchive.net/collection/TCGA-LUAD/">TCGA-LUAD</a></li> <li><a href="https://www.cancerimagingarchive.net/collection/TCGA-LUSC/">TCGA-LUSC</a></li> </ul> </li> <li>Links: <a href="https://doi.org/10.5281/zenodo.8290146">model weights 1, </a><a href="../record/8290169">model weights 2, </a><a href="https://github.com/bamf-health/aimi-lung-ct">github</a></li> </ul> <h4>lung2-ct.zip</h4> <ul> <li>Improved model version</li> <li>Segment Description: Lung and Nodules (3mm-30mm) from CT scans</li> <li>IDC Collections: <ul> <li><a href="https://www.cancerimagingarchive.net/collection/QIN-LUNG-CT">QIN-LUNG-CT</a>, <a href="https://www.cancerimagingarchive.net/collection/spie-aapm-lung-ct-challenge/">SPIE-AAPM Lung CT Challenge</a></li> </ul> </li> <li>Links: <a href="../records/11582738">model weights</a>, <a href="https://github.com/bamf-health/aimi-lung2-ct">github</a></li> </ul> <h4>lung-fdg-pet-ct.zip</h4> <ul> <li>Segment Description: Lungs and FDG-avid lesions in the lung from FDG PET/CT scans</li> <li>IDC Collections: <ul> <li><a href="https://www.cancerimagingarchive.net/collection/acrin-nsclc-fdg-pet/">ACRIN-NSCLC-FDG-PET</a></li> <li><a href="https://www.cancerimagingarchive.net/collection/anti-pd-1_lung/">Anti-PD-1-Lung</a></li> <li><a href="https://www.cancerimagingarchive.net/collection/lung-pet-ct-dx/">LUNG-PET-CT-Dx</a></li> <li><a href="https://www.cancerimagingarchive.net/collection/nsclc-radiogenomics/">NSCLC Radiogenomics</a></li> <li><a href="https://www.cancerimagingarchive.net/collection/rider-lung-pet-ct/">RIDER Lung PET-CT</a></li> <li><a href="https://www.cancerimagingarchive.net/collection/TCGA-LUAD/">TCGA-LUAD</a></li> <li><a href="https://www.cancerimagingarchive.net/collection/TCGA-LUSC/">TCGA-LUSC</a></li> </ul> </li> <li>Links: <a href="https://doi.org/10.5281/zenodo.8290054">model weights, </a><a href="https://github.com/bamf-health/aimi-lung-pet-ct">github</a></li> </ul> <h4>prostate-mr.zip</h4> <ul> <li>Segment Description: Prostate from T2 MRI scans</li> <li>IDC Collection: <a href="https://www.cancerimagingarchive.net/collection/ProstateX/">ProstateX,</a> <a href="https://www.cancerimagingarchive.net/collection/prostate-mri-us-biopsy/">Prostate-MRI-US-Biopsy</a></li> <li>Links: <a href="https://doi.org/10.5281/zenodo.8290092">model weights, </a><a href="https://github.com/bamf-health/aimi-prostate-mr">github</a></li> </ul>   Changelog <ul> <li>2.0.2 - Fix the brain-mr segmentations to be transformed correctly</li> <li>2.0.1 - added AIMI 2024 radiologist comments to qa-results.csv</li> <li>2.0.0 - added AIMI 2024 segmentations</li> <li>1.X - AIMI 2023 segmentations and reviewer scores</li> </ul>

opencc-by-4.0Nov 2023View details →

zenodo48/100

Segmenting magnetized plasma turbulence with aweSOM

This dataset contains a snapshot of a fully kinetic particle-in-cell simulation of freely evolving plasma turbulence, as described in <a href="https://iopscience.iop.org/article/10.3847/1538-4357/ac1c76" target="_blank" rel="noopener">Nättilä & Beloborodov (2021)</a>. This dataset was used in the analysis of <a href="https://arxiv.org/abs/2410.01878" target="_blank" rel="noopener">Ha et al. (2024)</a> and partially to develop <a href="https://github.com/tvh0021/aweSOM">aweSOM</a>. See the section: "Example : Intermittency detection in decaying plasma turbulence simulation" in the documentation of aweSOM for instructions on how to use these datasets.

opencc-by-4.0Oct 2024View details →

zenodo48/100

Hydrogen sulfide release via the ACE inhibitor Zofenopril prevents intimal hyperplasia in human vein segments and in a mouse model of carotid artery stenosis

The current strategies to reduce intimal hyperplasia (IH) principally rely on local drug delivery, in endovascular approach. The oral angiotensin converting enzyme inhibitor (ACEi) Zofenopril has additional effects compared to other non-sulfyhydrated ACEi to prevent intimal hyperplasia and restenosis. Given the number of patients treated with ACEi worldwide, these findings call for further prospective clinical trials to test the benefits of sulfhydrated ACEi over classic ACEi for the prevention of restenosis in hypertensive patients. Abstract Objectives Hypertension is a major risk factor for intimal hyperplasia (IH) and restenosis following vascular and endovascular interventions. Pre-clinical studies suggest that hydrogen sulfide (H2S), an endogenous gasotransmitter, limits restenosis. While there is no clinically available pure H2S releasing compound, the sulfhydryl-containing angiotensin-converting enzyme inhibitor Zofenopril is a source of H2S. Here, we hypothesized that Zofenopril, due to H2S release, would be superior to other non-sulfhydryl containing angiotensin converting enzyme inhibitor (ACEi), in reducing intimal hyperplasia in the context of hypertension. Materials Spontaneously hypertensive male Cx40 deleted mice (Cx40-/-) or WT littermates were randomly treated with Enalapril 20 mg (Mepha Pharma) or Zofenopril 30 mg (Mylan SA). Discarded human vein segments and primary human smooth muscle cells (SMC) were treated with the active compound Enalaprilat or Zofenoprilat. Methods IH was evaluated in mice 28 days after focal carotid artery stenosis surgery and in human vein segments cultured for 7 days ex vivo. Human primary smooth muscle cell (SMC) proliferation and migration were studied in vitro. Results Compared to control animals (intima/media thickness=2.3±0.33), Enalapril reduced IH in Cx40-/- hypertensive mice by 30% (1.7±0.35; p=0.037), while Zofenopril abrogated IH (0.4±0.16; p<.0015 vs. Ctrl and p>0.99 vs. sham-operated Cx40-/-mice). In WT normotensive mice, enalapril had no effect (0.9665±0.2 in control vs 1.140±0.27; p>.99), while Zofenopril also abrogated IH (0.1623±0.07, p<.008 vs. Ctrl and p>0.99 vs. sham-operated WT mice). Zofenoprilat, but not Enalaprilat, also prevented intimal hyperplasia in human veins segments ex vivo. The effect of Zofenopril on carotid and SMC correlated with reduced SMC proliferation and migration. Zofenoprilat inhibited the MAPK and mTOR pathways in SMC and human vein segments. Conclusion Zofenopril provides extra beneficial effects compared to non-sulfhydryl ACEi to reduce SMC proliferation and restenosis, even in normotensive animals. These findings may hold broad clinical implications for patients suffering from vascular occlusive diseases and hypertension.

opencc-by-4.0Dec 2020View details →

zenodo48/100

Klf14 mouse white adipose tissue histology DeepZoom files and AIDA annotations for visualisation of DeepCytometer white adipocyte segmentations

Latest description of this data set: <a href="https://github.com/MRC-Harwell/cytometer/blob/main/DATA.md">Data.md at cytometer project</a> <pre># Publications related to the data The data associated to the DeepCytometer project (https://github.com/MRC-Harwell/cytometer) is available from Zenodo (doi: 10.5281/zenodo.5137433 and 10.5281/zenodo.5149005). The histology and mouse measures were generated as part of the Small et al. 2018 study: > Small et al. "Regulatory variants at KLF14 influence type 2 diabetes risk via a female-specific effect on adipocyte size and body composition". Nature Genetics, 50:572–580, 2018. The hand traced data set, colour maps, and automatic segmentations were generated for the Casero et al. 2021 paper: > Casero et al. "Phenotyping of Klf14 mouse white adipose tissue enabled by whole slide segmentation with deep neural networks". bioRxiv, 2021. doi: [10.1101/2021.06.03.444997](https://www.biorxiv.org/content/10.1101/2021.06.03.444997v1.full). # Data protocols ## Histology and laboratory measures To develop and evaluate our methods we used Klf14tm1(KOMP)Vlcg C57BL/6NTac (B6NTac) mice tissue samples and additional data generated as part of the Small et al. 2018 study(Small et al. 2018). It should be noted that the single exon Klf14 gene is imprinted and only expressed from the maternally inherited allele(Parker-Katiraee et al. 2007). This was taken into account by (Small et al. 2018) by crossing a Het parent with a WT parent, so that each offspring inherited a WT allele from the WT parent, and the Klf14 gene knockout or a WT allele from the other parent (from the father, PAT, or the mother, MAT). We also take Klf14 imprinting into account by using as controls the PAT mice and comparing them to the MAT WT and MAT Het (or functional KO, FKO) mice.  We used a total of 76 Klf14-B6NTac mice (nfemale=nmale=38), of which 20 mice from the Control and FKO groups were used for training and testing the DeepCytometer pipeline, as well as the hand traced population experiment (summary in Table MICE). The histopathology screen involved fixing, processing and embedding in wax, sectioning and staining with Hematoxylin and Eosin (H&E) both inguinal subcutaneous and gonadal adipose depots. For paraffin-embedded sections, all samples were fixed in 10% neutral buffered formalin (Surgipath) for at least 48 hours at RT and processed using an Excelsior™ AS Tissue Processor (Thermo Scientific). Samples were embedded in molten paraffin wax and 8 μm sections were cut through the respective depots using a Finesse™ ME+ microtome (Thermo Scientific). Sampling was conducted at 2sxns per slide, 3 slides per depot block onto simultaneous charged slides, stained with haematoxylin Gill 3 and eosin (Thermo scientific) and scanned using an NDP NanoZoomer Digital pathology scanner (RS C10730 Series; Hamamatsu). Body weight (BW) and depot weight (DW) were measured with Satorius BAL7000 scales. ## White adipose tissue segmentation For cell area quantification, we applied DeepCytometer v8 to 75 inguinal subcutaneous and 72 gonadal whole histology slides with DeepCytometer (with the Corrected method), including the 20 slides sampled for the hand-traced data set, corresponding to 73 females and 74 males, to produce 2,560,067 subcutaneous and 2,467,686 gonadal cells (on average, 34,134 and 34,273 cells per slide, respectively). Full segmentation of all whole slides was performed with script [klf14_b6ntac_exp_0106_full_slide_pipeline_v8.py](https://github.com/MRC-Harwell/cytometer/blob/39358ed1d79df07d1d522b98728c7efd745513f7/scripts/klf14_b6ntac_exp_0106_full_slide_pipeline_v8.py). In this case, the segmentation contours were grouped by tiles in the output AIDA annotation `.json` file (one contour per cell, one file per slide). Non-white adipocyte contours were filtered out, and white adipocyte contours were aggregated into an AIDA annotation `.json` file with a single tile with script [klf14_b6ntac_exp_0106_annotations_postprocessing_v8.py](https://github.com/MRC-Harwell/cytometer/blob/39358ed1d79df07d1d522b98728c7efd745513f7/scripts/klf14_b6ntac_exp_0106_annotations_postprocessing_v8.py) (one contour per cell, one file per slide). # List of directories and files ## Casero et al. (2021) "DeepCytometer pipeline parameter files, Klf14 mouse white adipose tissue histology and hand-traced training contours" (doi: 10.5281/zenodo.5137433) ### `deepcytometer_pipeline_v8.zip` (60.6 MB) Weights, colourmaps, etc. necessary to run the pipeline (v8, with mode colour correction). This is the version of the pipeline described in the paper. There are 10 weight files per convolutional neural network (CNN), corresponding to 10-fold cross-validation * `klf14_b6ntac_exp_0086_cnn_dmap_model_fold_[0..9].h5`: Keras weights for the **EDT CNN** (Histology to Euclidean Distance Transform regression) * `klf14_b6ntac_exp_0089_cnn_segmentation_correction_overlapping_scaled_contours_model_fold_[0..9].h5`: Keras weights for the **Correction CNN** (Segmentation Correction regression) * `klf14_b6ntac_exp_0091_cnn_contour_after_dmap_model_fold_[0..9].h5`: Keras weights for the **Contour CNN** (EDT to Contour detection) * `klf14_b6ntac_exp_0095_cnn_tissue_classifier_fcn_model_fold_[0..9].h5`: Keras weights for the **Tissue CNN** (Pixel-wise tissue classifier) * `klf14_b6ntac_exp_0094_generate_extra_training_images.pickle`: training dataset description * **'file_list'**: list of SVG files with hand-traced contours for network training. Each SVG file has a corresponding TIFF file with the histology used for segmentation * **'idx_test'**: 10 lists with file indices for testing in 10-fold cross-validation * **'idx_train'**: 10 lists with file indices for training in 10-fold cross-validation * **'fold_seed'**: seed number used for the random number generator to assign file indices to folds * `klf14_b6ntac_exp_0098_filename_area2quantile.npz`: quantile colour maps calculated in `klf14_b6ntac_exp_0098_full_slide_size_analysis_v7.py` using the whole Klf14 data set with v7 of the pipeline, and used in earlier experiments, including some where v8 of the pipeline was used for segmentation. * `klf14_b6ntac_exp_0106_filename_area2quantile_v8.npz`: quantile colour maps calculated in `klf14_b6ntac_exp_0106_full_slide_pipeline_v8.py` using the whole Klf14 data set with v8 of the pipeline, and used in later experiments. * `klf14_training_colour_histogram.npz`: statistics from Klf14 histology images to be used in colour correction * **'xbins_edge'**, **'xbins'**: edges and centres of the bins used for histogram calculations * **'hist_r_q1'**, **'hist_r_q2'**, **'hist_r_q3'** * **'hist_g_q1'**, **'hist_g_q2'**, **'hist_g_q3'** * **'hist_b_q1'**, **'hist_b_q2'**, **'hist_b_q3'**: density quartiles (Q1, Q2, Q3) for RGB channels for each bin the histogram * **'mode_r'**, **'mode_g'**, **'mode_b'**: modes for RGB channels (this corresponds to the most typical background colour in the histology images) * **'mean_l'**, **'mean_a'**, **'mean_b'**: mean intensity for L*a*b channels of the image * **'std_l'**, **'std_a'**, **'std_b'**: intensity standard deviations for L*a*b channels of the image * `klf14_exp_0112_training_colour_histogram.npz`: other statistics from Klf14 histology images to be used in colour correction * **'p'**: vector of quantile values used in ECDF calculations * **'val_r_klf14'**, **'val_g_klf14'**, **'val_b_klf14'**: all intensity values for the RGB channels of Klf14 training images that contain at least a white adipocyte * **'f_ecdf_to_val_r_klf14'**, **'f_ecdf_to_val_g_klf14'**, **'f_ecdf_to_val_b_klf14'**: linear interpolation function that maps ECDF quantiles to intensity values in the Klf14 training data set. These functions can be used together with intensity->quantile interpolation functions calculated for a new histology image to perform histogram matching colour correction * **'mean_klf14'**, **'std_klf14'**: mean and standard deviation of the **'val_r_klf14'**, **'val_g_klf14'**, **'val_b_klf14'** vectors There are also weight files for the pipeline trained with all the data, instead of the 10-fold cross-validation partition. These were not used for the paper, but could be useful for future experiments * `klf14_b6ntac_exp_0101_cnn_dmap_model.h5`: Keras weights for the **EDT CNN** (Histology to Euclidean Distance Transform regression) * `klf14_b6ntac_exp_0104_cnn_segmentation_correction_overlapping_scaled_contours_model.h5`: Keras weights for the **Correction CNN** (Segmentation Correction regression) * `klf14_b6ntac_exp_0102_cnn_contour_after_dmap_model.h5`: Keras weights for the **Contour CNN** (EDT to Contour detection) * `klf14_b6ntac_exp_0103_cnn_tissue_classifier_fcn_model.h5`: Keras weights for the **Tissue CNN** (Pixel-wise tissue classifier) ### `histology.7z` (29.1 GB) 165 H&E histology whole slides from Hamamatsu scanner (`.ndpi`). ### `klf14.7z` (2.3 GB) Mice metadata, training/testing data sets for the pipeline, intermediate files created during training, and neural network weights for multiple experiments. * `klf14_b6ntac_meta_info.csv`: Klf14 mice metadata * **Animal Identifier**, **id:** unique ID for each mouse * **ko_parent:** heterozygous parent of origin for the KO allele (father, PAT or mother, MAT) * **sex:** female or male * **genotype:** wild type (KLF14-KO:WT) or heterozygous (KLF14-KO:Het) * **BW:** body weight (g) * **SC:** subcutaneous depot weight (g) * **gWAT:** gonadal depot weight (g) * **Liver:** livel weight (g) * **cull_age:** age at time of culling (days) * **BW_alive:** body weight measured before culling * **BW_alive_date:** age at time of BW_alive measure * **mother:** unique ID for mouse's mother * **mother_genotype:** mouse's mother genotype * `klf14_b6ntac_training`: Directory with hand-traced segmentations of training histology windows. 131 windows sampled from 20 whole slides, plus hand-traced contours that were used for training DeepCytometer and compute population distributions. These segmentations were used for CNN training, but note that there's a cleaned-up version of these data below, and it was the cleaned-up version that was used for the paper experiments * `ndpifile_row_YYYYYY_col_XXXXXX[.tif/.xcf/.svg]`: * **ndpifile:** name of the whole slide file (e.g. `KLF14-B6NTAC 36.1c PAT 98-16 C1 - 2016-02-11 10.45.00`) * **row_YYYYYY:** Y-coordinate of the top-left corner of the sampling window, in pixels * **col_XXXXXX:** X-coordinate of the top-left corner of the sampling window, in pixels * **.tif:** TIFF file with the histology sampling window * **.xcf:** Gimp file with the histology and hand-traced contours (the contours were drawn in Gimp) * **.svg:** SVG (Scalable Vector Graphics) that contains the hand-traced contours in the XCF file * `klf14_b6ntac_training_v2`: Same as `klf14_b6ntac_training`, but the hand-traced data set was cleaned up to remove small contours of dubious cells, or cells that are fully overlapped by others * `klf14_b6ntac_training_non_overlap`: Directory with intermediate images to train the networks. These images are generated by script [`klf14_b6ntac_training_non_overlap`](https://github.com/MRC-Harwell/cytometer/blob/main/scripts/klf14_b6ntac_exp_0077_generate_non_overlap_training_images.py) * `klf14_b6ntac_training_augmented`: Directory with intermediate images used to train the networks (using augmentation to reduce overfitting). These images are generated by script [`klf14_b6ntac_exp_0078_generate_augmented_training_images.py`](https://github.com/MRC-Harwell/cytometer/blob/main/scripts/klf14_b6ntac_exp_0078_generate_augmented_training_images.py) * `klf14_b6ntac_seg`: Deprecated. Directory to store whole slide coarse segmentations in old experiments (e.g. `klf14_b6ntac_exp_0076_generate_training_images.py`). Of little interest for most users * `klf14_b6ntac_results`: Deprecated. Directory to store miscellanea output from some experiments. Of little interest for most users ## Casero et al. (2021). "Klf14 mouse white adipose tissue histology DeepZoom files and AIDA annotations for visualisation of DeepCytometer white adipocyte segmentations" (doi: 10.5281/zenodo.5149005) ### `aida_data_Klf14_v8_images.7z` (16.9 GB) Histology images converted to DeepZoom so that they can be visualised with [AIDA](https://github.com/alanaberdeen/AIDA). To use this, decompress this file and put the resulting `images` directory in your `AIDA/dist/data/` directory. ### `aida_data_Klf14_v8_annotations.7z` (18 GB) White adipocyte segmentations in AIDA annotation `.json` files (one contour per cell, one file per whole slide). Each slide has the following files: * `SLIDENAME.json`: Soft link to the annotations file that we want to associate to slide `SLIDENAME.ndpi`, e.g. `SLIDENAME` = `KLF14-B6NTAC-PAT-39.2d 454-16 B1 - 2016-03-17 12.16.06` * `SLIDENAME.lock`: Empty file used to tell the pipeline that `SLIDENAME.ndpi` has already been processed or is being currently processed * `SLIDENAME_coarse_mask.npz`: File with the coarse tissue segmentation of `SLIDENAME.ndpi` and the internal state of the pipeline (execution times, steps, etc) * `SLIDENAME_exp_0106_auto.json`: Annotations (all segmentations without filtering from the Auto algorithm, i.e. segmentation without object overlap). Contours are grouped by the tile they were processed in * `SLIDENAME_exp_0106_auto_aggregated.json`: Filtered annotations (non-white adipocytes removed) of the Auto algorithm. All contours aggregated into a single tile * `SLIDENAME_exp_0106_corrected.json`: Annotations (all segmentations without filtering from the Corrected algorithm, i.e. segmentation with object overlap). Contours are grouped by the tile they were processed in * `SLIDENAME_exp_0106_corrected_aggregated.json`: Filtered annotations (non-white adipocytes removed) of the Corrected algorithm. All contours aggregated into a single tile To use this, decompress this file and put the resulting `annotations` directory in your `AIDA/dist/data/` directory.</pre>

opencc-by-4.0Jul 2021View details →

zenodo48/100

OpenForensics: Multi-Face Forgery Detection And Segmentation In-The-Wild Dataset [V.1.0.0]

OpenForensics is the first large-scale dataset posing a high level of challenges. This dataset is designed with face-wise rich annotations explicitly for face forgery detection and segmentation. With its rich annotations, OpenForensics dataset has great potentials for research in both deepfake prevention and general human face detection. Project Page:  https://sites.google.com/view/ltnghia/research/openforensics

opencc-by-4.0Oct 2021View details →

zenodo48/100

Dataset for Semantic Segmentation of Fishing Trajectories

This is the dataset that was manually labelled by the author during his research work for the paper "Semantic Segmentation of AIS Trajectories for Detecting Complete Fishing Activities" in MDM 2022.

opencc-by-4.0Nov 2022View details →

zenodo48/100

International benchmark for ALS individual tree segmentation

This upload aims to provide an international benchmark dataset for airborne LiDAR-based individual tree segmentation algorithm comparison and development.

opencc-by-4.0Dec 2022View details →

zenodo48/100

Doodleverse/Segmentation Zoo Res-UNet model for NOAA ERI/4-class segmentation of RGB 512x512 images

This Residual-UNet model is trained on 1,179 pairs of human-generated segmentation labels and images from Emergency Response Imagery (ERI) collected by US National Oceanic and Atmospheric Administration (NOAA) after Hurricane Barry, Delta, Dorian, Florence, Ida, Laura, Michael, Sally, Zeta, and Tropical Storm Gordon. The dataset is available here: https://doi.org/10.5281/zenodo.7268082 Models have been created using Segmentation Gym: Code - https://github.com/Doodleverse/segmentation_gym Paper - https://doi.org/10.1029/2022EA002332   The model takes input images that are 512 x 512 x 3 pixels, and the output is 512 x 512 x 4, corresponding to 4 classes: <ol> <li>water</li> <li>bare sediment</li> <li>vegetation</li> <li>development (roads, buildings, power lines, parking lots, etc.)</li> </ol>   Included here are 6 files with the same root name: <ol> <li> '.json' config file: this is the file that was used by Segmentation Gym to create the weights file. It contains instructions for how to make the model and the data it used, as well as instructions for how to use the model for prediction.</li> <li>'.h5' weights file: this is the file that was created by the Segmentation Gym function `train_model.py`. It contains the trained model's parameter weights. It can called by the Segmentation Gym function `seg_images_in_folder.py`.</li> <li> '_model_history.npz' model training history file: this numpy archive file contains numpy arrays describing the training and validation losses and metrics. It is created by the Segmentation Gym function `train_model.py`</li> <li> '.png' model training loss and mean IoU plot: this png file contains plots of training and validation losses and mean IoU scores during model training. A subset of data inside the .npz file. It is created by the Segmentation Gym function `train_model.py`</li> <li>'.zip' of the model in the Tensorflow ‘saved model’ format. It is created by the Segmentation Gym function `utils/gen_saved_model.py`</li> <li>'_modelcard.json' model card file: this is a json file containing fields that collectively describe the model origins, training choices, and dataset that the model is based upon. There is some redundancy between this file and the `config` file (described above) that contains the instructions for the model training and implementation. The model card file is not used by the program but is important metadata so it is important to keep with the other files that collectively make the model and is such is considered part of the model</li> </ol> Additionally, BEST_MODEL.txt contains the name of the model with the best validation loss and mean IoU

opencc-by-4.0Feb 2023View details →

zenodo48/100

Segmentation maps of giant Antarctic icebergs

Segmentation maps of the giant Antarctic icebergs B30, B31, B34, B35, B41, B42 and C34 derived with a U-net approach, Otsu thresholding and k-means. Individual images are roughly one month apart. A description of the method and discussion of results can be found here: Braakmann-Folgmann, A., Shepherd, A., Hogg, D., and Redmond, E.: Mapping the extent of giant Antarctic icebergs with Deep Learning, EGUsphere [preprint], https://doi.org/10.5194/egusphere-2023-858, 2023. Please cite this paper when using or refering to the data.

opencc-by-4.0Apr 2023View details →

zenodo48/100

IMC Segmentation Pipeline results of example IMC data

If you are working with these files, please cite them as follows: Windhager, J., Zanotelli, V.R.T., Schulz, D. et al. An end-to-end workflow for multiplexed image processing and analysis. Nat Protoc (2023). <a href="https://doi.org/10.1038/s41596-023-00881-0">https://doi.org/10.1038/s41596-023-00881-0</a>This repository hosts the results of processing example imaging mass cytometry (IMC) data hosted at <a href="http://10.5281/zenodo.5949116">10.5281/zenodo.5949116</a> using the IMC Segmentation Pipeline available at <a href="https://github.com/BodenmillerGroup/ImcSegmentationPipeline">https://github.com/BodenmillerGroup/ImcSegmentationPipeline</a> (DOI: <a href="http://10.5281/zenodo.6402666">10.5281/zenodo.6402666</a>) v3.6. Please refer to <a href="https://github.com/BodenmillerGroup/steinbock">https://github.com/BodenmillerGroup/steinbock</a> as alternative processing framework and <a href="http://10.5281/zenodo.6043600">10.5281/zenodo.6043600</a> for the data generated by steinbock.The following files are part of the analysis.zip folder when running the IMC Segmentation Pipeline:<ul><li>cpinp: contains input files for the segmentation pipeline</li><li>cpout: contains all final output files of the pipeline: cell.csv containing the single-cell features; Experiment.csv containing CellProfiler metadata; Image.csv containing acquisition metadata; Object relationships.csv containing an edge list indicating interacting cells; panel.csv containing channel information; var_cell.csv containing cell feature information; var_Image.csv containing acquisition feature information; images containing the hot pixel filtered multi-channel images and the channel order; masks containing the segmentation masks; probabilities containing the pixel probabilities.</li><li>histocat: contains single channel .tiff files per acquisition for upload to histoCAT (<a href="https://bodenmillergroup.github.io/histoCAT/">https://bodenmillergroup.github.io/histoCAT/</a>)</li><li>crops: contains upscaled image crops in .h5 format for ilastik (<a href="https://www.ilastik.org/">https://www.ilastik.org/</a>) training</li><li>ometiff: contains .ome.tiff files per acquisition, .png files per panorama and additional metadata files per slide</li><li>ilastik: multi channel images for ilastik pixel classification (_ilastik.full) and their channel order (_ilastik.csv); upscaled multi channel images for ilastik pixel prediction (_ilastik_s2.h5); upscaled 3 channel images containing ilastik pixel probabilities (_ilastik_s2_Probabilities.tiff).</li></ul>The remaining files are part of the root directory:<ul><li>docs.zip: Documentation of the pipeline in markdown format</li><li>IMCWorkflow.ilp: Ilastik pixel classifier pre-trained on the example data</li><li>resources.zip: The CellProfiler pipelines and CellProfiler plugins used for the analysis</li><li>scripts.zip: Python notebooks used for pre-processing and downloading the example data</li><li>src.zip: Scripts for the imcsegpipe python package</li></ul>

opencc-by-4.0Feb 2022View details →

zenodo48/100

Pre-training with simulated ultrasound images for breast mass segmentation and classification - dataset

Dataset assosiated with the MICCAI Workshop on Data Engineering in Medical Imaging paper: "Pre-training with Simulated Ultrasound Images for Breast Mass Segmentation and Classification"

opencc-by-4.0Oct 2023View details →

ScienceDex guides

Understand access before you commit

These curated guides explain access requirements, typical timelines, costs, and reuse considerations for widely used research datasets.

Compare curated datasets

Allen Brain Atlas

Allen Brain Atlas is an Allen Institute collection of brain map atlases, datasets, APIs, and analysis tools covering mouse, human, and non-human primate brain resources.

allen-brain-atlas

neuroscienceopenDocumentation, web resources, and API references are available online.

Last verified 2026-04-30Open record

Annotated Behaviour and Observability Dataset (ABODe)

ABODe is a University of Edinburgh DataShare dataset for behavior classification in group-housed mice using home-cage video, identities, bounding boxes, ground-plate positions, and annotator labels.

abode-home-cage

behavioral-neuroscienceopenThe DataShare record exposes download links for annotations, documentation, license text, and the zipped per-snippet data directory.

Last verified 2026-04-30Open record

DANDI Archive for NWB datasets

DANDI is a BRAIN Initiative archive for publishing and sharing neurophysiology data, including electrophysiology, optophysiology, and behavioral data packaged as NWB and related standards.

dandi-nwb

electrophysiologyopenPublished Dandiset metadata and archive endpoints are available through the production DANDI API.

Last verified 2026-04-30Open record

International Brain Laboratory public data

The International Brain Laboratory public data releases expose standardized mouse decision-making experiments, including Neuropixels recordings, widefield calcium imaging, behavior, and session metadata accessed through the ONE API.

ibl

behavioral-neuroscienceopenPublic sessions can be searched and loaded from the IBL public data server through ONE.

Last verified 2026-04-29Open record

OpenNeuro

OpenNeuro is a free, open platform for sharing neuroimaging datasets, with public search, dataset pages, and download paths for web, S3, DataLad, and the OpenNeuro CLI.

openneuro

neuroscienceopenPublished datasets are available on demand over the internet.

Last verified 2026-04-29Open record